{ "device": "NVIDIA GB10", "cutlass_dsl": "4.6.2", "metadata": { "width": 1344, "height": 768, "frames": 124, "steps": 12, "sampler_step": 1, "seed": 440420, "text_tokens": 100, "hidden_shape": [ 37810, 5376 ], "segments": [ [ 0, 100, 1 ], [ 100, 514, 2 ], [ 514, 37810, 0 ] ] }, "linear": "attn_out_proj", "mnk": [ 2048, 5376, 7168 ], "full_activation_rows": 37810, "modeled_chunk_count": 19, "ring": { "row_capacity": 2048, "producer": "vortex_native_quantize_nvfp4", "qdata_bytes": 7340032, "sfa_bytes": 917504, "logical_bytes": 8257536 }, "parity": { "tensor_scale_equal": true, "fp4_difference_count": 0, "block_scale_difference_count": 0, "output_equal": true, "max_abs": 0.0, "mean_abs": 0.0, "comfy_output_equal": true }, "timing": { "warmup": 10, "iterations": 100, "producer_ms": 0.22272863388061523, "global_scale_ms": 3.323794250488281, "cute_consumer_ms": 0.775920639038086, "modeled_cute_chunk_ms": 0.9986492729187012, "modeled_comfy_chunk_ms": 1.0811997032165528, "actual_into_ring_cute_gemm_ms": 1.008663330078125, "comfy_consumer_ms": 0.8584710693359375, "actual_into_ring_comfy_gemm_ms": 1.1376866912841797, "reference_vortex_scale_comfy_pack_gemm_ms": 1.178064956665039, "modeled_canonical_reference_ms": 25.70702842712402, "modeled_canonical_cute_ring_ms": 22.488397521972654, "modeled_canonical_comfy_ring_ms": 24.939841384887693, "actual_ring_vs_reference": 0.8562034923214508, "comfy_ring_vs_reference": 0.9657249244598827, "modeled_canonical_cute_ring_vs_reference": 0.8747956842123643, "modeled_canonical_comfy_ring_vs_reference": 0.9701565256983631 } }